Clear memory options

[dbuscombe-usgs]

This PR ...

1. adds two new config options:

• CLEAR_MEMORY [default=False] will clear the keras session and carry out garbage collection at the end of each training epoch. This is for memory leaks that have occurred on large datasets
• LOAD_DATA_WITH_CPU [default=False] will use the CPU to do the tf.data pipeline, which may also help memory leaks or other memory issues

2. run_eagerly=True is removed (it is not needed)

3. adds parallel processing option do_parallel to the seg_images_in_folder script, which applies models to imagery in parallel using joblib

4. added a new/updated conda yml and recipe for windows and linux

[dbuscombe-usgs]

This PR includes fixes for https://github.com/Doodleverse/doodleverse_utils/issues/27 and https://github.com/Doodleverse/segmentation_gym/issues/117

Added options to the wiki https://github.com/Doodleverse/segmentation_gym/wiki/04_Creation-of-%60config%60-files

[dbuscombe-usgs]

I made a few more minor edits I have made on the checkpoints branch. These deal with

1. situations when the save_weights does not make the "fullmodel" version of the weights, which is now dealt with by a try/except loop because it happens apparently randomly (probably a bug in the tensorflow build), and it is not crucial
2. added the 'profile' to the seg_images_in_folder script to modify outputs
3. played with joblib's Parallel method in seg_images_in_folder script - it works sometimes, but randomly errors out almost every time. It is an unreproducible bug. It is left in as a hidden (experimental) method while we figure out a better alternative

Please review this soon. I'd like to get it incorporated so it is up to date with the rest of the doodleverse, which is now using doodleverse version 29 https://github.com/Doodleverse/doodleverse_utils/releases/tag/v0.0.29